The Operator's
Prompt Vault

I want to find underserved SaaS opportunities. My background is [your background/skills]. My target audience is [target audience, e.g. "e-commerce brands under $1M ARR"]. Search my audience's pain through this lens: 1. What manual processes are they still doing in spreadsheets that should be automated? 2. What do they pay for in agencies/freelancers that software could replace? 3. What data do they have that they're not acting on because it's too hard to synthesize? For each pain, output: - Problem statement (one sentence) - Current workaround (how they're solving it today) - Why a SaaS could beat the workaround - Rough willingness-to-pay signal - One customer quote that might appear in a Reddit/forum post about this pain Give me 10 pains. Then rank them by: (a) frequency, (b) urgency, and (c) money already being spent on bad solutions.COPY

I'm exploring a SaaS idea: [describe your idea in 2-3 sentences]. Help me do a realistic, bottoms-up market size estimate. Not TAM/SAM/SOM fluff — I need founder-grade thinking. 1. Who are the atomic buyers? (Identify the exact job title/persona that swipes the card) 2. How many of them exist globally? Break it into: Tier 1 English-speaking markets first. 3. What's a realistic price point based on: (a) value delivered, (b) comp tools, (c) buyer budget category? 4. What's a realistic penetration rate for a bootstrapped SaaS in year 1, 2, 3? 5. What's my year-1 revenue ceiling if I execute well but not perfectly? Be honest. Poke holes in the assumptions. Tell me if this looks like a lifestyle business or a real scalable opportunity.COPY

I want to automate this business workflow: [describe the workflow end-to-end]. Break it down as an AI systems architect would: 1. Map every manual step currently required 2. Classify each step as: (A) Fully automatable now with existing AI tools, (B) Partially automatable with human-in-the-loop, (C) Requires human judgment — do not automate yet 3. For each Type A step, name the specific tool or API that handles it (e.g. Claude, GPT-4o, Make.com, Zapier, Apify, Airtable, etc.) 4. Design the trigger → process → output chain for the full automated version 5. Identify the #1 failure point in this automation and how to build a fallback Output this as a workflow diagram in text (using arrows and labels) that I can share with a developer or VA.COPY

I want to build an AI agent system that handles: [describe the agent's purpose and scope]. Design the agent architecture: 1. Agent type: Single agent, multi-agent with orchestrator, or pipeline chain? 2. Memory design: What does this agent need to remember between sessions? (Classify: short-term context, long-term user data, external knowledge base) 3. Tool list: What tools/APIs does this agent need access to? List them. 4. Decision logic: Write pseudocode for the core decision tree this agent follows 5. Human escalation triggers: When should this agent stop and ask a human? 6. Output spec: What does a successful run produce? (File, message, database update, webhook, etc.) Also: What's the cheapest viable stack to run this agent in production for under $50/month?COPY